论文部分内容阅读
以多幂次变差的测量为理论基础,考虑到有限样本规模的局限以及市场微观结构噪声的影响,提出交错取样门限多幂次变差方法并将其用于中国股市高频已实现波动的细分,区分出连续波动与跳跃波动。根据已实现波动、连续波动与跳跃波动的不同统计特征,分别为已实现波动与连续波动建立LHAR-V-CJ模型,为跳跃波动强度建立LHAR-SJ-C模型,为跳跃波动间隔时间建立LACH-DJ-C模型,引入异质非对称性。使用沪深300指数的实证表明,已实现波动及连续波动与跳跃强度、跳跃间隔时间呈现出不同的非对称性特征,且本文提出的各非对称性模型较现有模型均有较明显的拟合能力改进。
Taking the measurement of multiple power variation as the theoretical basis and considering the limitations of the finite sample size and the influence of the market microstructure noise, a multi-order variation method of interleaved sampling threshold is proposed and applied to the fluctuation of the high frequency of the Chinese stock market Subdivision, distinguish between continuous fluctuations and jump fluctuations. According to the different statistical characteristics of the volatility, continuous fluctuations and jumps, the LHAR-V-CJ model is established for the realized fluctuations and the continuous fluctuations, the LHAR-SJ-C model is established for the jump fluctuation strength, and the LACH -DJ-C model, introducing heterogeneous asymmetry. Empirical tests using CSI300 show that the volatility, continuous fluctuation and jump strength have been realized, and the jumping interval time shows different asymmetry characteristics. And the asymmetric models proposed in this paper have more obvious than the existing models Ability to improve.